Title: Probabilistic Description of Traffic Breakdowns Caused by On-ramp Flow

Abstract: The characteristic features of traffic breakdown near on-ramp are analyzed. To describe this phenomenon the probabilistic description regarding the jam emergence as the formation of a large car cluster on highway inside the synchronized traffic is constructed. In these terms the breakdown occurs through the formation of a certain critical nucleus in the metastable vehicle flow, which is located near the on-ramp. The strong cooperative car interaction in the synchronized traffic enables us to treat the size of critical jam nuclei as a large value and to apply to an effective one-lane model. This model assumes the following. First, the growth of a car cluster is governed by the attachment of cars to the cluster whose rate is mainly specified by the total traffic flow. Second, the cluster dissolution is determined by the car escape from the cluster whose rate depends on the cluster size directly. Third, the generation of one-car clusters (preclusters) is caused by cars entering the main road from the on-ramp. The appropriate master equation for the car cluster evolution is written and the generation rate of critical jam nuclei is found. The obtained results are in agreement with the empirical facts that the characteristic time scale of the breakdown phenomenon is about or greater than one minute and the traffic flow rate interval inside which traffic breakdowns are observed is sufficiently wide. Besides, as a new results, it is shown that the traffic breakdown probability can be analyzed, at least approximately, based solely on the data of the total vehicle flow without separating it into the vehicle streams on the main road and on-ramp when the relative on-ramp flow volume exceeds 10%--20%.